CHAPTER 12

Becoming an Analytical Healthcare Organization

Business intelligence and analytics excellence “is achieved when organizations have in place the strategy, people, process, and technology approaches that result in business impact, value, and effectiveness.”1 Analytics excellence, as it relates to healthcare quality improvement (QI), is when the strategies, people, processes, and technologies are applied to improvement initiatives and positively impact the quality and performance of a healthcare organization (HCO).

Being an analytical organization requires more than simply acquiring or possessing the tools and technology of analytics. To become an analytical organization, an HCO must ensure that:

Items 1 through 4 in the list above are issues that must be addressed in the HCO’s analytics strategy. (See Chapter 3 for a discussion of analytics strategies that drive healthcare improvement.) Ultimately, the analytics strategy is responsible for ensuring that the organization’s analytical, business, and technology requirements are in alignment and that efforts on all fronts are focused on achieving the same goals.

Item 5 above represents a gap between analytics development and the use of analytics within the HCO, and is the most challenging step to becoming an analytical organization. Several of the reasons this is a challenge include the following barriers:

• Resistance to change. Healthcare leaders and decision makers are understandably very busy, and some may feel that the “old way” of decision making (replete with incomplete information and “gut feelings”) is just fine. To overcome resistance to change, the value of analytics (in terms of making more effective decisions in less time) must be demonstrated by clear, tangible results.
• Rapid business change. Sometimes by the time an analytical tool or report is built and deployed, the precipitating crisis has passed and the HCO has moved on to other issues. To prevent this situation, analytics teams must be agile and able to respond quickly to the evolving needs of the organization. Understanding and focusing on the overall quality and performance goals of the organization (communicated via the analytics strategy) also helps analytics teams to be better prepared for the kinds of analytics the HCO will be requiring.
• Distrust. Perhaps the most frustrating cause of a gap in analytics utilization is distrust of the information. All it takes is two reports to show different numbers for the supposedly same metric, and executives and quality teams will be suspicious. This distrust can be rectified through strong data governance and precise data and indicator definitions. Addressing this also needs vigilance on the part of analytics teams and data stewards to ensure that all potential sources of information are in alignment and all possible sources of disparity in data are known, monitored, and mitigated when required.

The barriers identified above prevent full usability of analytics throughout the HCO, and each manifests some gap in analytics capability, whether in knowledge, skill, understanding, or technology. To further address these barriers, there are three main areas of excellence on which analytics teams must focus to achieve excellence:

Of the three layers of analytics excellence above, I consider execution the most important, because that is the layer that addresses the analytics usability gap. In fact, I believe that HCOs should relentlessly pursue execution excellence (that is, actually use analytics for real-world problems), because doing so will drive the attainment of technical and professional excellence. Without execution and use of analytics throughout the organization, there will be no demand for technical and professional excellence. In extreme cases, senior management may be oblivious to the analytics needs and capabilities of the organization, thereby resulting in substandard and/or haphazardly deployed information management technologies being rolled out and poorly resourced and coordinated analytics teams left to fend for themselves. Organizations that focus too much on technical development of analytics infrastructure at the expense of execution, however, may run the risk of building the “ultimate” and likely very costly analytics solution that nobody in the organization can actually use because the focus on what actually was required by the organization was lost.

You do need to ensure that you always use the highest-quality data possible, and that the technical infrastructure is as robust as possible. In reality, however, no technical solution is ever going to be “perfect” (i.e., there will always be some data quality problems, and the technology doesn’t always work). Furthermore, no organization is going to have the exact right mix of skills. But these issues can be overcome and evolve as analytics is used within the HCO. To ensure that your HCO does not become permanently stuck in a “build” phase, find ways to use the capabilities that already exist (and stretch those a little bit)—and continue to meet the needs of the organization while providing clear examples of how analytics demonstrates value for the organization.

Requirements to Become an Analytical Organization

Healthcare is in a state of change. HCOs can be responsible for driving necessary changes within to achieve business, quality, and performance goals, or they can be perpetually reacting to the pressures around them without really knowing the best action to take. Analytics is one of the fundamental tools that enable HCOs to achieve change.

Most of this book has focused on very tangible, quantifiable things such as data, how to manipulate data, how to demonstrate a change in performance using data, and how to turn data into information and insight that is useful to healthcare leaders and QI teams. But many HCOs have terabytes or more of data and multiple dashboards, and are making decisions, but do not seem to be achieving their performance and quality goals. Having data and dashboards alone aren’t sufficient for becoming an analytical organization and achieving excellence in healthcare quality and performance. To become an analytical HCO requires:

• Strategy
• Leadership and commitment
• Focus
• Agility
• Teamwork

Strategy

As discussed at length in Chapter 3, the purpose of the analytics strategy is to guide the HCO’s ability to rapidly respond to the information needs of key decision makers while maintaining a consistent direction in supporting the quality and business goals of the HCO. A solid analytics strategy will help enable the analytics team to become a strategic information resource for business improvement and not simply a purveyor of reports and data.

An analytics strategy is the starting point to help organizations achieve maximum benefit from analytics. A completed strategy will help an organization identify what it does well, what it needs to do better, where it can consolidate, and where it needs to invest. The analytics strategy should not be set in stone either; it needs to evolve as the analytics needs of the organization and its stakeholders evolve, as technology becomes better and/or less expensive, and as the state of the art in analytics itself changes. An organization should not be afraid to revisit the strategy frequently to ensure that it is up to date and that the execution of the strategy is successfully meeting all stated requirements.

Leadership and Commitment

Leadership is often the deciding factor in an analytical organization; it is, after all, the leaders within an HCO who “have a strong influence on culture and can mobilize people, money, and time to help push for more analytical decision making.”2 Although pursuing healthcare improvement through the use of QI initiatives coupled with information and insight generated from analytics seems like it would be a “slam dunk,” HCOs are often resistant to that kind of change. It then takes strong leadership within the HCO to begin or continue down that road and to overcome resistance.

Key analytics-related responsibilities of leaders within the HCO are to:

• Keep the HCO focused on strategic goals and objectives. It is a primary job of leaders within the HCO to ensure that all QI activities align with the strategic goals of the organization, and that analytics is in alignment with those goals. Leaders also need to know when it is necessary to deviate from those stated strategic goals when unexpected, pressing issues arise. When those issues are resolved, it falls on leadership to reorient any efforts that may have been diverted back to working on strategic goals. This is where the value of quality and analytics strategies demonstrates its worth.
• Promote and champion use of analytics throughout the HCO. Support decision making using analytics. Support the deployment and use of analytics tools. Recognize and support the need for enterprise data structures to enhance analytics.
• Enforce data governance policies and procedures. Data management, regardless of the scope and scale of the data being managed, cannot be placed on autopilot. Accurate analytics requires that all aspects of data (including the process definitions from which data is derived) are constantly and consistently managed. Not all leaders will be part of governance efforts, but all should recognize and value the need for strong data governance in ensuring that high-quality data is available for reporting and analytics, and that accurate results and meaningful insights depend on having high-quality data. Effective leaders will also recognize the need for agility, and will not enforce any more layers of approval than absolutely necessary to ensure the quality of data.
• Encourage, enable, and reward innovation and experimentation. At the heart of healthcare improvement lies innovation. Innovation involves finding newer (and presumably better) solutions to existing problems. For example, changing processes and adding communications technology that result in decreased turnaround time for hospital inpatient beds would be an innovation. Applying analytics to solve pressing QI problems within an HCO is also innovative. HCOs that achieve their quality and performance goals usually have well-established cultures of innovation, where it is permitted and even expected that healthcare staff seek out innovative ways to improve pressing quality and performance issues. (Of course, these innovations are always being evaluated to ensure they are having the desired effect on processes and outcomes.)
• Provide analytics teams required training and tools. Leaders must ensure that analytics teams have the proper tools and training to perform their work effectively, and must do their best to protect the time of analytical teams to focus on work that is aligned with the strategic focus and needs of the HCO.

Focus

Commitment and leadership are both necessary to enable analytics teams to focus on building what is important to the organization and necessary for achieving the organization’s quality and performance goals. I like to classify the type of work that analytics teams do into three categories:

Strategic activities are vitally important to the HCO, as these help to build a sustainable analytics infrastructure. Examples of sustainable analytics infrastructure include organized and intuitive analytics portals that enable self-serve access to information and insight when people require it. Overall, strategic activities are those that create a sustainable, accessible information resource that helps to identify and direct action toward organizational information needs.

Tactical activities, on the other hand, are those that are in support of actual QI initiatives. These typically involve preparing baseline data, developing project-specific dashboards, and evaluating process outcomes in detail. Tactical-type activities are an extension of the strategic activities, except they are directed at providing the information and insight that individual QI teams require.

Finally, reactionary activities are those in response to an urgent request. These types of requests can be very distracting to a team, depending on the scope of work required and how quickly it is required. These activities may be related to simple data requests, or required because of a critical incident or similar circumstance within the HCO.

It is my observation that many of the strategic-level activities that should be a priority for an analytics team get sidelined for reactionary activities, which is ironic, given that many strategic-level activities can actually improve the tools and systems available to enable designated individuals to access the information they require. One of the critical roles that leadership and commitment play, then, is allowing analytics teams to focus on strategic objectives that will both reduce the number of reactionary-type activities and enable analytics teams to participate more often on quality and performance improvement projects.

Agility

Analytics professionals are very highly skilled, solution-oriented, and motivated individuals coming from a variety of backgrounds, including computer science, engineering, statistics, and epidemiology (to add my own educational background!). The effective development, implementation, and use of analytics can be resource-intensive, involving an in-demand small group of individuals, specialized tools, and unique knowledge and skill sets.

The need for and perceived value of analytics within HCOs is increasing, and there are many different projects that compete for the same analytics skills and resources. Regardless of the size of the “analytics shop” within an HCO, whether it’s a handful of analysts within a department or program or a large team within a business intelligence competency center, it will not take them very long to get bogged down in the minutiae of day-to-day data, report, information, and application requests. Within operations of a healthcare environment, it’s the biggest fires, the loudest voice, or the proverbial squeaky wheel that gets the attention of resources. Unfortunately, the squeaky wheels are not necessarily the priorities that are truly important to the organization as a whole, QI in particular, or even the analytical teams themselves.

The challenge, then, is how to exactly determine what is important and should be getting the attention of the analytics team. It may be tough for analytics teams to know if they have the right tools and resources to do the jobs asked of them, and it is difficult to know what jobs to do from the realm of competing priorities. This is where an analytics strategy is necessary. The analytics strategy is essential for helping to sort and prioritize incoming requests for information.

Healthcare QI needs to be agile—that is, it must be able to respond to issues and requests as they arise. QI projects are no longer years-long efforts; time frames to achieve expected results are now measured in days and weeks. The development of analytics to address quality issues cannot become a barrier to the rapid initiation of QI projects. That is why analytics teams must understand the needs of QI teams (and in fact should work side by side).

Analytics teams must know how to take raw data and present it in a form that is quickly usable by the QI teams, and QI teams must know how to ask for information in ways that the analytics teams can respond to. It doesn’t matter which types of frameworks are guiding QI efforts—Lean, Six Sigma, and others require the analytics teams and QI teams to be on the same page to bring usable analytics to the front lines.

It is unlikely that an HCO will be starting from scratch—that there are no existing QI teams and projects, and no business intelligence, analytics, or report-development resources. What is likely, however, is that the QI and analytics teams do not work closely together. In most organizations, QI teams must follow “report request” (or similarly outmoded) processes just to submit a request for a report, dashboard, or other information. QI initiatives can be highly energizing and exciting events, especially when participating in rapid improvement events or other similar activities. Nothing stifles this excitement, or otherwise inhibits innovation, more than not having the right information to make decisions or to intelligently identify issues. Even worse is when team members must go through obtuse data request procedures simply to obtain data.

When process changes are being made and evaluated in a span of hours or a few days, waiting weeks for data and other analytics is simply unacceptable. This is why I strongly advocate for analytics team members to be part of QI initiatives, or at least for there to be very strong connections between the QI and analytics teams. QI teams must know whom to talk to for the data, information, and analysis that they need. In return, the analytics team must be both aware that such improvement initiatives are happening and prepared to provide as rapid turnaround as possible. This is where a well-defined quality strategy and strong executive support for analytics is necessary, to establish and support these tight connections so that the analytics required for QI projects is available when required, and not only at the convenience of the analytics team. The need for this agility is why analytics teams cannot be encumbered with numerous data requests that detract from their ability to respond to initiatives of strategic and tactical importance.

Building Effective Analytical Teams

Throughout my career, I have seen many different types of people, with many different backgrounds, excel in healthcare analytics. I believe that it is the strong diversity of backgrounds and skills that analytics professionals possess that makes analytics indispensable for healthcare quality and performance improvement initiatives.

There are an abundance of opinions highlighting various qualities and attributes of data scientists, business intelligence professionals, and analysts. Much of the discussion, however, has centered around the math, data, or technology skills of analytics professionals. Because my focus is on the application of analytics for quality and performance improvement, the qualities I view as ideal for analytics professionals involved in these activities typically are situated within the intersection of IT, the business, and the QI activities of the HCO.

With this in mind, several of the traits I view as important for healthcare analytics professionals are as follows:

• Natural curiosity. As more healthcare data becomes available via the proliferation of electronic health records, there is much to be learned about the data available and in turn much to be learned from what the data tells us. Healthcare analytics professionals should be naturally curious and revel in asking “what” and “why,” realizing that these questions do not expose ignorance but are truly the only way to gain full understanding of a problem.
• Innovative mind-set. Healthcare quality and performance improvement initiatives require a great deal of innovation to identify more efficient and effective workflows and processes. To help achieve the required levels of innovation, healthcare analytics professionals must see analytics not as “report development,” but as a way to building the “information tools” necessary to solve pressing healthcare issues. They are willing, able, and excited to leverage all the technology and information available to maximum extent (whether it’s experimenting and adopting new visualizations or trying novel analytical approaches). They strive for effective yet creative solutions that provide efficient access to the right information to the right people when it is needed.
• Business focus. Improving healthcare quality and performance requires a strong and thorough understanding of processes and workflows. Analytics to support QI initiatives must align with and provide insight into the business of providing care. This is why healthcare analytics professionals must focus on the business, striving to know the pertinent details of the healthcare domains in which they work. After all, it is these details of the business that add the necessary context to data that helps it become “information” and “insight.”
• Technological savvy. In many ways, analytics operates at the heart of healthcare information technology, given that analytical solutions typically integrate data from multiple data sources (such as clinical and financial systems). Many systems and steps are involved in getting data from source systems into a location and format available for effective analysis. Having said that, however, experienced healthcare analytics professionals don’t need to be tech jockeys (that is, they don’t need to be hardcore programmers or serious database administrators). But they should be comfortable and proficient with the current and emerging technologies, such as business intelligence platforms and data cleaning, analysis, and visualization tools. This means being comfortable in using more than just a spreadsheet.
• Team player. Effective healthcare analytics projects depend upon having effective analytics teams. This means working well with other members of healthcare analytics and QI teams, all while respecting the differing points of view that professionals in other disciplines (such as nurses, physicians, and laboratory technologists) bring to the discussion. It also means communicating well; healthcare analytics professionals must both listen to and understand what others are saying, and articulately convey their own opinions and knowledge to others who may not be analytics experts.

Healthcare QI is now a multidisciplinary effort, involving a range of experts including clinical, administrative, technology, and process engineering professionals. Due to the different roles and teams in which healthcare analytics professionals may find themselves, a strong mix of technical, interpersonal, and analytical skills is essential to successfully operate in today’s challenging healthcare environment.

Summary

Every HCO is unique and faces different challenges based on factors ranging from its patient population and their healthcare requirements to funding limitations, legislation pressures, and the makeup of clinical and administrative staff. Healthcare quality and performance improvement requires a wide range of changes, from reducing and eliminating waste and inefficiencies to analyzing processes in detail and engineering new solutions to improve patient outcomes. HCOs may begin with solving issues related to poor flow and advance to more complex patient safety and clinical outcomes issues.

HCOs that achieve their goals do so by allowing their staff to try out new and innovative ideas, to evaluate those ideas within mini-experiments, and to implement and deploy those innovations that are demonstrated to improve the way healthcare is delivered and HCOs are managed. Those same organizations utilize and rely on two of their most strategic assets—their healthcare data and the people who create insights from that data—to provide evidence-based guidance for individual improvement initiatives from inception to completion. This is the way to healthcare transformation.

Yes, healthcare QI initiatives can exist and be successful without the benefit of analytics. But analytics makes those projects much more efficient and effective. Likewise, analytics does not need to be integrated into structured QI methodologies to have a dramatic impact on operational and clinical decision making. But organizations that are striving to improve healthcare to achieve improved outcomes are more likely to succeed once their QI initiatives are fully able to leverage analytics assets and capabilities. The powerful insights possible with analytics combined with a structured approach to identifying, implementing, and evaluating improvement opportunities can greatly improve the likelihood that QI activities can achieve changes that matter and outcomes that last.

Notes

1. John Boyer et al., Business Intelligence Strategy: A Practical Guide for Achieving BI Excellence (Ketchum, ID: MC Press, 2010), 7.

2. Thomas H. Davenport, Jeanne G. Harris, and Robert Morison, Analytics at Work: Smarter Decisions, Better Results (Boston: Harvard Business School Publishing, 2010), 57.